Grok API relay services 26 in total

Relay services that support the Grok family, ordered by overall score. Each service's uptime, response latency and measured model count come from real probing, and every page states its source and method.

About Grok relay services

Far fewer services support Grok than the families above — a relatively niche category with genuine demand. The narrow coverage comes from a limited number of upstream channels, which means only a handful of services can supply it reliably.

A small field cuts two ways. The upside is that choosing is simple; the downside is that unit prices are usually less competitive than other families — without much competition there is little room to negotiate. So the question here is availability rather than cheapness.

With so few options, a more conservative approach makes sense: test on a small amount first, confirm that calls are reliable before committing more, and keep a backup channel configured so you are not depending on a single provider.

How to choose within this group

The picks below are computed live from the current measurements, not fixed conclusions — change the data and the picks change with it. Each states its basis, and you can open the detail page to check it yourself.

  • Most stable:Highest current uptime. Where samples are thin we show the public-dataset baseline instead, and the detail page states the source.
  • Fastest:Home-page response time from the most recent probe. A single measurement is noisy, so treat it as a range rather than an absolute.
  • Broadest model coverage:This uses the model count our endpoint probe actually detected, not the support list the service advertises — the two are measured differently, and the advertised figure is usually the more optimistic one.

Key parameters side by side

The top 10 services by overall score, for comparison. The "models" column is the count our own probe detected, not the support list the service advertises.

ServiceRegion UptimeLatencyModels PaymentAPI
Yuegle API
api.yuegle.com
China direct 100% 603 ms — — Has /v1/models
api.gpt.ge
api.gpt.ge
China direct 100% 450 ms — — Has /v1/models
api.vveai.com
api.vveai.com
China direct 100% 412 ms — — Has /v1/models
便携AI聚合API
api.bianxieai.com
China direct 98.36% 786 ms — — Has /v1/models
API易
docs.apiyi.com
China direct 100% 789 ms — Alipay · WeChat Pay Unconfirmed
MegaLLM
megallm.io
Global / overseas 98.36% 1060 ms 62 Credit card Unconfirmed
PoloAPI-直连官转中转API
poloapi.com
China direct 100% 955 ms — WeChat Pay Unconfirmed
gpt.ge
gpt.ge
China direct 100% 662 ms — WeChat Pay Unconfirmed
APIYI
apiyi.com
Global / overseas 100% 1269 ms — — Unconfirmed
Poixe AI
poixe.com
China direct 100% 804 ms 178 — Has /v1/models

Above are the top 10 by score; this family has 26 in total — see them all in the full filtered list.

Official price baseline

A relay's "rate" only means something against a baseline, and this section is it. Prices come from vendor pricing pages, at list price, excluding volume discounts.

ModelVendor InputOutput CacheContext
Grok 4.3
grok-4.3
xAI $1.25 $2.5 — 1000K
Grok 4.20
grok-4.20
xAI $1.25 $2.5 — 1000K
Grok 4.7
grok-4.7
xAI $1.6 $4.8 $0.4 500K
Grok 4.6
grok-4.6
xAI $2 $6 $0.5 500K
Grok 4.1 Fast
grok-4.1-fast
xAI $0.2 $0.5 — 2000K

Units: USD per million tokens. The complete vendor price list is in the model price table; to turn a rate into real spending, use the calculator in the rankings.

Frequently asked questions

The questions we see most often about the Grok family. The guidance applies to the family as a whole, not to any single service.

Why do so few services support Grok?

Mainly because upstream channels are limited. Compared with mature ecosystems like GPT and Claude, there are noticeably fewer ways to connect to Grok, so fewer services can supply it reliably. It also means there is little point comparing prices aggressively — there is not much to choose from.

How stable is Grok through relays?

Overall uptime figures are on our comparison board. With so few services available, treat the stability tier as the main criterion rather than the unit price alone.

How do I check whether a service really offers Grok?

Request /v1/models and look for the matching model identifier. Our probe records what each service actually returns, which you can verify on its detail page.

Browse by model family

You can also switch dimension and browse by use case, the rankings or the uptime monitor.